ISSS608 2016-17 T1 Assign3 Linda Teo Kwee Ang Analysis

Introduction

Approach

Analysis

Responses

Evaluation

The main purpose of this assignment is to understand how people move and communicate in the park, as well as how patterns changes and evolve over time, and what can be understood about motivations for changing patterns. By examining the movement data, with special attention paid on the outliers, we are able to surface some visitors of interest, and could draw hypotheses on what could had happened, especially on the eventful day. Apparently some crimes had taken place on 8 Jun 2014 (Sun), including a case of vandalism. The analysis below would seek to shed some light on these crimes.

Communication data is critical to drawing inferences on the hypotheses. The movement data, while was sufficient enough to highlight suspicious movement, would not be able to draw links between visitors and groups. Hence, the communication data would come in to verify the links, and provide more insights to what could had happened.

ANALYSIS

A total of 11,374 unique IDs, representing 11,374 visitors over the 3 days.

• 62% of the visitors had only visited the park on 1 day: Mostly on Sunday (51%), followed by Sat (29%) then Fri (20%).
• 22% of the visitors had visited the park over 2 days: vast majority (86%) were over Sat to Sun.
• 16% of the visitors (or 1,813 persons) went to the park on all 3 days.

Opening hours of Creighton Pavilion

It was noted that there would not be any check-ins into Creighton Pavilion from 9:30am to 11:30pm and from 2:30pm to 4:30pm. It is possible that the Pavilion would be closed from visitors during Scott Jone’s show at Grinosaurus Stage.

Visitors of interest based on short park visits

The visitors had spent duration ranging from 00:00:00 to 14:52:31 (hh:mm:ss) per visit (assuming they did not exit from the park in-between). There does not appear to be much differences between the 3 days. The following visitors were flagged up due to missing records or extreme short visits:

• 2 rows of visitor record on 8 Jun (Sun) at 1:14:03 PM did not contain any visitor ID nor X, Y coordinates. These 2 rows may be erroneous.
• Other visitors who spent extremely short duration at the park may also be flagged up for investigations. The following list had duration of less than 4 hours during at least one of their visits. Of interest, a group of 8 visitors had spent similar length of time on all 3 days.

Visitor ID	6 Jun (Fri)	7 Jun (Sat)	8 Jun (Sun)
1781128	2:33:51	-	-
1763672	3:10:50	-	-
657863	5:41:35	0:09:36	-
1187304	-	1:28:08	-
1748887	-	2:48:02
1680161	-	3:05:58	-
898576	-	11:50:57	2:02:23
47441	-	-	2:32:53
1269018	-	-	2:50:12
921888	-	-	3:11:54
1983765	10:52:19	12:07:39	3:31:46
1787551	8:28:41	8:28:30	3:28:32
1080969	8:28:53	8:28:30	3:28:33
1600469	8:28:53	8:28:41	3:28:33
1935406	8:28:53	8:28:52	3:28:33
1781070	8:28:42	8:28:52	3:28:34
1629516	8:28:42	8:28:41	3:28:43
644885	8:28:42	8:28:42	3:28:44
521750	8:28:42	8:28:52	3:28:55

Special Group of 8

Visualisation of the groups’ movements were done using Tableau. This group of 8 was noted to check-in from East Entrance at Kiddie Land (point 99,77), then travelled in a group. Their movements around the park is consistent throughout the 3 days:

https://www.youtube.com/watch?v=iC19VwzPI1Q

• Check-in at East Entrance (Kiddie Land point 99,77) at around 8:45am
• Walked one round through Tundra Land, Wet Land and Coaster Alley, and reached Grinosaurus Stage at about 9:30am, then exited from Grinosaurus Stage at 11:30am and went back the same path, exiting the park at around 12:15pm.
• Check-in at East Entrance again at about 1:45pm, took the same path and reached Grinosaurus Stage at about 2:30pm. The group then exited at about 4:30pm, and walked back the same path, exiting the park at about 5:15pm.
• The above was consistent on both 6 Jun (Fri) and 7 Jun (Sat), but on 8 Jun (Sun), the group did not return after exiting at about 12:15pm.
• There were no check-ins into any of the attractions and rides.

There are a few hypotheses for this group of 8:

• Hypothesis 1: This group may be employees of the park who were specifically involved in 2 daily performances at Grinosaurus Stage, from 9:30am to 11:30am and 2:30pm to 4:30pm. However it was not stated whether employees of the park would need to use the app. It is also unlikely there is any need for them to use the app since its purpose is more for visitors.

• Hypothesis 2: This group is solely interested in watching the show at Grinosaurus Stage, hence they had only gone to Grinosaurus Stage for both morning and afternoon shows. However it is quite unlikely that this group would watch five consecutive shows. Furthermore they could have taken a shorter route to get to the Grinosaurus Stage from East Entrance.

• Hypothesis 3: This group is related to the Amazing Scott Soccer Showcase Show, which was scheduled for 2 daily performances from 6 to 8 Jun 2014. From the description of the assignment question, the stage shows by Scott Jones would take place twice everyday from 6 to 8 Jun, and in addition, a show of memorabilia related to his illustrious career would be displayed in the park’s Pavilion. This is likely to mean that Scott Jone’s shows would be held at Grinosaurus Stage, and a separate exhibition was held at Creighton Pavilion. This hypothesis is very likely, that this is the entourage for Scott Jones, possibly including his manager and security guards. Matched with the movement records of the group of 8, it is hence deduced that Scott Jone’s shows took place from 9:30am to 11:30am, and then from 2:30pm to 4:30pm. The show was supposed to be held twice daily, but the last show on 8 Jun 2014 (Sun) did not take place. It is very likely that the afternoon show on 8 Jun 2014 (Sun) was cancelled due to the crime.

The other 5 with short duration on 8 Jun 2014 (Sun)

Emphasis is placed on 8 Jun 2014 (Sun) since it is likely to be the eventful day. There are 5 others who had short total duration in the park on 8 Jun:

• 898576 - this visitor strangely ceased movements around 10:20am, when he was at point 43,56 Kauf’s Lost Canyon Escape. As he/she had entered the park at 8:17am, he had only stayed for about 2 hours on records. Either he was lost in the Lost Canyon (and maybe still there today), otherwise there might be problems in tracking his movement via the app. The latter is more likely, as his communication data showed that he was still communicating at 10:13pm on 8 Jun 2014 (Sun). It is not known whether he was aware of the problems, or whether he had deliberately meddled with the app to hide his tracks. It is unlikely that he had lost his phone and someone else picked it up), because his contacts were similar before and after 10:20am.

• 47441 - this visitor had entered the park at 9:40am via East Entrance. There were no communications made throughout this period, until 11:52am when he made external communications. From his entry, it was noted that he was moving quite fast ((4 seconds per grid compared to norm of 6 seconds), and did not check into any rides/attractions. Instead, he stayed around some of the rides, for between 12 and 30 minutes. He went over to Creighton Pavilion at around 11:57am, and stayed there for almost 10 minutes, then he went to the East Entrance and leaving the park. This visitor was likely to be alone, and did not have any communication records with any other visitors in the park Instead, he made a few external communications, with the first one at 11:52am when he was near to Smoky Wood BBQ. When he reached Creighton Pavilion at 11:57am, he made 4 more external messages. During his 10 minutes at Creighton Pavilion, he made several calls to 839736, which is believed to the assistance/helpdesk. It is hypothesized that 47441 was badly affected by the vandalism, and was trying to alert the park management through the assistance/helpdesk. He probably felt upset after seeing the vandalism, hence decided to leave the park.
• 1269018 - this visitor was probably running. he was moving across the grid at 3 seconds per 15 metres. He appeared to be travelling alone, and did not check-into any rides, only moved around the vicinity. He left within 3 hours of entry. There was no communication made by him.

• 921888 - this is another "running" visitor, who moved as fast as 1269018. However both of them did not seem to be travelling together. There was very low communication data for 921888, only a few external communications, and with 839736. While 921888 appeared to be a normal visitor who went for the rides, he had ‘’’stayed outside Creighton Pavilion for 1 hour and 3 minutes at 9:06am’’’. Notably, this was the time when Creighton Pavilion was closed from public, hence there was no reason why he would wait there for 1 entire hour.

• 1983765 - this particular visitor visited the park on all 3 days, and his movement records seemed to be erroneous. Furthermore, he was walking very slowly, at half of the speed of a common visitor. It is hypothesized that he could be surveying the park to plot for his crime, if he was the perpetrator. There was however, no communication data on this visitor.

  - on 7 Jun 2014 (Sat) 3:06pm when he was noted to be at 2 places at the same time.

  - on 7 Jun 2014 (Sat) 3:45pm, he was at Creighton Pavilion, but the next instance at 4pm, his record bounced to Grinosaurus Stage. There seems to be missing movement records in-between.

  - on 7 Jun 2014 (Sat) 8:18pm, he was at 2 places at the same time - outside Creighton Pavilion (at point 31,34) and also near to Ichthyoroberts Rapids (at point 76,36). They appeared to be 2 separate persons who then moved towards point 35,63, and then they left the park together at around 8:36pm via West Entrance.

  - on 8 Jun 2014 (Sun), he entered the park at 8:15am, went direct and reached Creighton Pavilion at 8:32am, then left the Creighton Pavilion at 9:08am to go to Scholtz Express and stayed there for 2 hours and 19 minutes, then left the park immediately at 11:47am via East Entrance.

Visitors of interest based on long duration at a single point

Generally a visitor would take 6 seconds to move from one XY coordinate to another. Hence the time difference between each movement record for the same visitor ID is usually 6 seconds. The speed of walking is roughly at 15 metres per 6 seconds. When a visitor spends more time at the same location, it may mean that he/she is using the facilities of that location, eg dining if it is an eating place, or the visitor is waiting. For example, at 8:17am on 7 Jun 2014, visitor ID 100025 spent 37 minutes at point 69,72, which is actually a shopping premise Staskosaurus Designs, hence this visitor was shopping. On average, a visitor would spend 46 minutes and 50 seconds on a single XY point if not on the move. After slight adjustments to remove duration spent at the 3 entrance points (as visitors could actually exit and re-enter the park, thus the duration upon their exit should not be counted as duration in the park), the average duration at each point was 46 minutes and 34 seconds, with mode at 1 hour and 3 minutes.

Using a Graph Builder in JMP, we are able to compare the duration of visitors at each XY point, and pick out the outliers. Then these individual visitors' movement records were viewed using Tableau.

• On 8 Jun 2014 (Sun), 3 visitors (1502920, 416790 and 461004) spent more than 3 hours at point 32,33 which is Creighton Pavilion. The 3 of them were likely to be travelling together. The strange part about these 3, is that they remained inside Creighton Pavilion on 8 Jun (Sun) even after it was closed from visitors at 9:30am. In fact they stayed for at least 2 hours inside the pavilion during that period. They left the pavilion together at 11:46am, and continued to stay in the park (and checked into some attractions/rides together) until 9:40pm.

• On 8 Jun 2014 (Sun), 1 visitor bearing 1983765 spent 2 hours and 19 minutes at point 6,43, which is Scholtz Express. This same visitor was also flagged up earlier for short total duration in the park on 8 Jun 2014, and was walking very slowly. It may be noted that this visitor's movement records might be erroneous. Hence even though it was stated that he was at Scholtz Express, there is a likelihood that he might be somewhere else.

• On 8 Jun 2014 (Sun), 6 visitors (227221, 392618, 1336607, 1483705, 1722376 and 2063022) were noted to spend more than 5 hours per person in the restrooms. In particular, they had all gone to Tyrannosaurus Rest at about 2:55pm after going to Grinosaurus Stage, and stayed at Tyrannosaurus Rest for 5 hours and 48 minutes. To stay at the restroom for such long time in a group, there is a likelihood that some/all person(s) in the group might be seeking medical attention at the restroom. After the stay at the restroom, the group of 6 left the park together via Main Entrance at about 8:44pm.

• On 6 Jun 2014 (Fri), 4 visitors (1486047, 1690685, 1797150 and 629048) spent at least 6 hours at point 76,22 which is Grinosaurus Stage. In fact, they had check-into Grinosaurus Stage twice, at around 9:51am and 1pm. They were likely to be travelling together. The peculiar part is that while all 4 travelled together on 6 Jun (Fri), only 1690685 entered the park on 7 and 8 Jun 2014.

• On 6 Jun 2014 (Fri), 5 visitors (103006, 1412235, 1937834, 313073 and 657863) spent almost 5 hours at point 21,33 which is Alvarez Beer Garden. One of them even stayed overnight at the park. This group had entered the park together via Main Entrance at 4:11pm, and went straight to Chensational Sweets. After spending about 35 minutes there, they proceeded to Alvarez Beer Garden, reaching at about 5:04pm. They remained at Beer Garden until 9:50pm, then moved towards Scholtz Express and stayed for about 27 minutes there. After that 4 of them left via Main Entrance. The 5th visitor, 657863, appeared to remain at Scholtz Express overnight, probably drunk and only woke up at about 8am on 7 Jun (Sat), and swiftly left the park via Main Entrance.

• On 6 Jun 2014 (Fri), 2 visitors (174727 and 193621) spent almost 4 hours at point 47,79 which is Granite Slab Pizza. One of them visited the park for all 3 days.

Slow-walkers

The most suspicious visitor is 1983765 who had erroneous movement records and suspiciously-slow pace of walking. A check on the other visitors revealed that there were also 2 other visitors who fit the same description – walked slowly and were likely to be alone. These 2 visitors were ‘’’1089132’’’ and ‘’’1723967’’’. Similar to 1983765, both 1089132 and 1723967 also did not have any communication records.

Visitor ID	Walking speed	Dates of visit	Entry/Exit timings
1983765	13 seconds per grid Used West Entrance	6 Jun (Fri) 7 Jun (Sat) 8 Jun (Sun)	– entered at 9:28am, left 8:21pm – entered at 8:28am, left 8:36pm – entered at 8:15am, left 11:47am
1089132	16 seconds per grid Used Main Entrance	6 Jun (Fri) 7 Jun (Sat) 8 Jun (Sun)	– entered at 9:17am, left 7:13pm - entered at 9:01am, left 7:56pm – entered at 8:06am, left 7:22pm
1723967	18 seconds per grid Used East Entrance	6 Jun (Fri) 7 Jun (Sat) 8 Jun (Sun)	– entered at 9:32am, left 7:47pm – entered at 8:53am, left 10:15pm – entered at 9:01am, left 2:51pm

Linking them up

Given that there were 3 suspicious visitors (1502920, 416790 and 461004) who stayed inside Creighton Pavilion during the period of offence, and 3 possible “slow-walkers” (1983765, 1089132 and 1723967) believed to be performing surveillance at the park on all 3 days. Given that 1983765’s movement records seemed to be erroneous, there is a possibility that he had disrupted the app data to disguise his movements. Since he was able to do that, probably he would also do likewise for his accomplices. Hypothetically, the app IDs 1502920, 416790 and 461004 could also be misused by this trio to disguise their real identities?

Other suspicious visitors: 1269018, 921888, but there was insufficient communications data to work on them.